Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.
Sources: National Vulnerability Database, CISA KEV, Exploit-DB
CVE-2026-109195 Overview
CVE-2026-109195 is an authorization flaw in Red Hat Quay, a container image registry. The vulnerability allows an authenticated attacker to retrieve repository auto-prune policy configurations belonging to repositories in other namespaces. The flaw stems from missing namespace scoping when the application resolves policy details by repository name and policy identifier. Exploitation requires the attacker to control a repository that shares the target's name and to know the target policy's unique identifier. The weakness is classified under CWE-639: Authorization Bypass Through User-Controlled Key.
Critical Impact
Authenticated users in one namespace can disclose auto-prune policy configuration details from unrelated namespaces when they can guess or obtain policy identifiers and reuse repository names.
Affected Products
- Red Hat Quay (container image registry)
- Repository auto-prune policy API endpoints
- Multi-tenant Quay deployments with shared repository names across namespaces
Discovery Timeline
- 2026-10-12 - CVE-2026-109195 published to the National Vulnerability Database (NVD)
- 2026-10-15 - Last updated in NVD database
Technical Details for CVE-2026-109195
Vulnerability Analysis
Quay exposes auto-prune policies that administrators configure to automatically remove old or unused repository tags. The policy lookup logic retrieves a policy by combining a repository name with a policy identifier. The lookup does not validate that the resolved policy belongs to the namespace the requesting user controls. An authenticated attacker who owns a repository with the same name as a victim's repository can therefore request a policy identifier scoped to the victim and receive the response. The disclosed data is limited to auto-prune policy configuration, resulting in a confidentiality-only impact without integrity or availability consequences.
Root Cause
The root cause is improper enforcement of namespace boundaries during object lookup, a classic Insecure Direct Object Reference pattern ([CWE-639]). The backend treats the policy identifier as a user-controlled key and trusts it without cross-referencing the authenticated principal's namespace scope. Repository name collisions between namespaces, which Quay permits by design, make the authorization gap reachable.
Attack Vector
Exploitation occurs over the network against an authenticated session. The attacker must satisfy two preconditions: owning a repository whose name matches a victim repository, and possessing the unique identifier of the target policy. The combination of required knowledge and setup corresponds to high attack complexity. No user interaction is required on the victim side. The vulnerability discloses auto-prune policy details only; it does not permit modification of policies or disruption of registry operations.
No verified proof-of-concept code is published. See the Red Hat CVE-2026-109195 Advisory and Red Hat Bug Report #2549646 for vendor-provided technical details.
Detection Methods for CVE-2026-109195
Indicators of Compromise
- Requests to Quay auto-prune policy API endpoints that reference policy identifiers not previously created within the requesting user's namespace.
- Repeated API calls enumerating sequential or guessed policy identifiers against a repository name shared across namespaces.
- Audit log entries showing successful policy reads from accounts that have never configured auto-prune policies.
Detection Strategies
- Correlate Quay application logs with namespace ownership metadata to identify policy reads crossing tenant boundaries.
- Alert on API clients requesting auto-prune policy details for repository names that appear in multiple namespaces.
- Baseline normal auto-prune API usage per account and flag deviations, especially bursts of lookup requests.
Monitoring Recommendations
- Enable verbose Quay audit logging for all /api/v1/repository/*/autoprunepolicy/ endpoints and forward events to a central analytics platform.
- Monitor authentication logs for accounts that create many short-lived repositories whose names mirror those of other tenants.
- Review Quay access patterns regularly for authenticated reconnaissance against policy identifier spaces.
How to Mitigate CVE-2026-109195
Immediate Actions Required
- Apply the fixed Quay release referenced in the Red Hat CVE-2026-109195 Advisory as soon as it is available for your deployment channel.
- Audit existing auto-prune policies and treat their identifiers as sensitive values that should not be shared outside the owning namespace.
- Review Quay audit logs for prior access to auto-prune policy endpoints by accounts outside the owning namespace.
Patch Information
Red Hat tracks remediation in Red Hat Bug Report #2549646 and publishes fixed package versions through the Red Hat CVE-2026-109195 Advisory. Administrators should consult the advisory for the exact Quay version that enforces namespace scoping on auto-prune policy lookups.
Workarounds
- Restrict Quay user registration and repository creation to trusted principals to limit the pool of potential attackers who can mint colliding repository names.
- Rotate or recreate sensitive auto-prune policies after patching so that any previously exposed policy identifiers are invalidated.
- Reduce exposure of policy identifiers by excluding them from screenshots, support tickets, and chat logs shared across tenants.
# Configuration example
# Review auto-prune policies per namespace via the Quay API after authentication
curl -H "Authorization: Bearer ${QUAY_TOKEN}" \
https://quay.example.com/api/v1/repository/${NAMESPACE}/${REPO}/autoprunepolicy/
Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.
Sources: National Vulnerability Database, CISA KEV, Exploit-DB

